The fixed effect of longitudinal standard bingeing on network connectivity are shown in Table 2. Only the fixed effect of longitudinal standard bingeing on DMN-VAN connectivity survived the FDR correction (β=−0.01, t(168.05)=−3.44, puncorrected<.001, FDR=.03). Greater standard bingeing was associated with a negative change (Time 2 < Time 1) in DMN-VAN connectivity (Figure 1), controlling for the initial binge group memberships, longitudinal network changes, in-scanner movements, scanner effects, SES, sex, and age. That is, at higher standard bingeing levels, DMN-VAN connectivity at Time 1 is higher than Time 2. The inverse pattern is true at lower standard bingeing levels. Furthermore, in separate models controlling for a) substance use prior to session 1 (non-binge drinking, standard/extreme bingeing, marijuana and tobacco use), b) longitudinal substance use (non-binge drinking, marijuana and tobacco use), and c) substance use within 24 hours of scanning (alcohol, marijuana, tobacco use, caffeine and prescription medication) in addition to the covariates in the original models, the negative association between change in DMN-VAN and longitudinal standard bingeing remained largely unchanged (FDRa=.03, FDRb=.06, FDRc=.04).
